The Great Big NBN Sticky Thread

EDIT by TheWedgie:
To keep all the NBN questions and discussion in one space, as a lot of it has been asked and answered before, after speaking with Akh-Horus and a few others, I have merged a few of the main NBN related threads into this one and stickied it.

I will be working on rules at some point, but in the short term this is for TECHNICAL discussion around the NBN, so keep politics out of it as nobody wins those arguments.
We're lucky to have a few people with inside knowledge of the NBN (either from NBNco or ISP's) that are willing to share their knowledge and experience here, so don't abuse that, or expect miracles or special favours.



------------------Akh-Horus' original post
Saw a chap out the front measuring the property line and putting details in to an iPad. Wandered up and asked what he is doing - turns out I will have a NBN node placed right at the front of my house in 6 to 12 months. Distance to where the house is less than 20 meters.

Other peoples distances? Will be interesting.

Also nothing to say the node may not be moved but this appears to be where the plan says it will be.

 

The distance is based on the pillar (the grey dome thing), not the node. They node is normally placed very close to the pillar - is your house close to that?

From WP, Generally, you should be able to get the full 100Mbps if your cable length is 500m or less. for me, I am unlucky and sync at ~45Mbps despite being 550m cable length from the pillar. I found out that there are many joints to side streets which affects my max attainable rate :/

 

I think my record so far is just under 800m from node to pillar. it's as close as it can get, but as far away as it needs to be.

mmm...

joints have relatively little effect on DSL per se. what they are is an opportunity for a bad joint, or a gauge or a construction change that can decrease performance, but all of those could happen in 1 joint too. more joints just increases the chance of bad luck.

it is a mistake to try to estimate DSL performance on the basis of distance, especially when the actual cable length cannot be determined with any accuracy. there are just too many factors that can affect performance beyond simple distance. nor does a "good" ADSL2+ service mean VDSL2 will work well on the same line.

 

nek minnut

 

We had an NBNCo rep in our consultancy a few weeks ago, turns out our new designs are to use type 6 pits and installing their FDH gear in the pits now... (usually used type 5/8 for distribution and 9 pits next to a pillar style FDH)

 

The pillar and node are both at the end of my street, so less than 500m. NBNCo are forecasting RFS in Feb '18. I'm not convinced it's much of a value proposition though. I'm potentially going to pay more for roughly the same speed I have now. But at least I can go faster if I want to.

 

perhaps as a matter of interest....


Click to view full size!

to understand in context:

• this is based on real world testing (not lab)
• you don't know what gauge or construction your line is, or whether it changes
• you don't know what the spectral profile on the DSLAM is
• you don't know if your line has any faults or correctable transmission defects affecting performance, like excess crosstalk, bridge taps, split pairs or electrical issues
• this is for an optimised combination of DSLAM and modem chipset - YMMV considerably with anything else

typical street copper runs about 37 to 53dB per kilometre depending on cable variant, but you also need to factor in jointing loss, plus make allowance for lead-in and internal cabling, and the tie cabling from the pillar to the DSLAM.

so realistically, at 500m I would expect a line in good repair to return around 75Mbps bitrate.

 

that is the 3rd generation "skinny fibre" for FTTP and for MTM node backhaul, not for FTTP as such. the existing copper stays where it is.

 

Oh wow, that's brutal. Moved the entire cabinet just to shut him up lol

 

Not sure if that was meant for me, but I'm currently showing 41dB indicating a line length of approx 3.5km, and I average a realistic figure of around 10Mbps. I'll initially sign on at 12/1 because I don't want to pay any more.

 

No. They decided there was a better location for the cabinet.

 

I'm aware of that site. the relocation was not done due to the one individual, to my knowledge it made the mains power connection easier. the premises in question is still on wireless because it's outside the coverage of the pillar the node is tied to, and that's not going to change unless the resident chooses to pay for a bespoke connection under the technology switch programme.

 

it doesn't really matter why they moved it.

it only matters that it makes for a brilliant BOFH type story.

 

you cannot make such direct comparisons. it doesn't work that way.

for starters, your modem shows measured signal attenuation, which is an average across all of the DMT subtones carrying payload. the graph I posted it at a fixed frequency of 3.75MHz, which is far higher than ADSL2+ can possibly reach. so you would need to make a realistic estimate of the average frequency your connection is running at (assuming the modem displays spectral load), know what attenuation per kilometre that frequency translates to to estimate line length, then apply that to the graph (at something like 40dB/km) to derive a realistic estimated performance under VDSL2.

and that's assuming your current DSLAM port doesn't have an attenuation pad, which will distort the calculations, plus I can pretty much guarantee the NBN DSLAM will have a pad configured (but you won't know the value of either), and that you don't have a short bridge tap anywhere which is quite likely to impact the VDSL2 bitrates significantly, whereas it may not affect ADSL2+ at all.

finally, the NBN DSLAM probably isn't located at the same place along the line as the existing ADSL DSLAM, unless it's in a Telstra cabinet, and even then you would need to know the differential tie cable lengths if they are dissimilar to any degree.

you just can't make an accurate guess. there are too many engineering factors that have significant impacts.

that would be a wise idea, and see what attainable rates the line can support before considering a higher speed plan.

 

I accept that until it's actually connected, I just wont know. It can't be any slower, surely!!

The termites keep eating the power cables under the street. 4 times in 13 years they've dug a hole to repair them. I'm not sure why they haven't eaten the phone cables yet?!

 

matters to me, because the implication of the story is that it was moved to spite the resident, which was not the case. the network build doesn't revolve around individuals, but engineering standards.

but if anyone wants BOFH, I'm happy to change their location service class to satellite only.

 

Fair enough, but still shut him up though.

 

What sort of costs are people looking at under the tech switch program?

 

5 figures, at least. Have a read on wp

 

Just did one at work @ 30k

10k of that is for tunnel bore work but still 20k for 0.7km fibre is what we were quoted